j o u r n a l o f t r a f fi c a n d t r a n s p o r t a t i o n e n g i n e e r i n g ( e n g l i s h e d i t i o n ) 2 0 1 8 ; 5 ( 5 ) : 3 6 1 e3 7 1

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/jtte

Original Research Paper

Guidelines for roundabout circulatory and entry widths

based on vehicle dynamics

Maksym Diachuk*, Said M. Easa

Department of Civil Engineering, Ryerson University, Toronto, ON M5B 2M2, Canada

highlights

A universal articulated vehicle's 2D steerability dynamic model is developed.

The speed mode of long-wheelbase vehicle affects the swept path on a roundabout.

The entry roadway width is estimated using vehicle dynamics simulation model.

article info abstract

Article history: Current guidelines for roundabout circulatory roadway width are based on a static method
Received 29 September 2017 that does not consider circulatory speed. In addition, the roundabout entry width is based
Received in revised form on practical experience. This paper presents a method for determining the circulatory and
4 January 2018 entry widths based on a two-dimensional vehicle dynamics model that involves a system
Accepted 11 January 2018 of differential equations of curvilinear motion. The method considers the interactions
Available online 1 August 2018 between a vehicle and road geometric elements, including tire sideslip, vehicle weight,
vehicle speed, and vehicle stability. Three design vehicles are considered: intermediate
Keywords: semitrailer (WB-12), interstate semitrailer (WB-20), and city transit bus (CITY-BUS). Design
Roundabout guidelines for the required circulatory width are established for different circulatory
Circulatory roadway width speeds (0e60 km/h) and different inscribed circle diameters (30e80 m). To simplify the
Vehicle dynamics guidelines, for each design vehicle and for each inscribed circle diameter (ICD) the
Inscribed circle diameter regression model of circulatory roadway width as a power function of circulatory speed
was fitted. Guidelines for entry width were also established for typical conditions. The
results show the efficiency of the proposed method which provides smaller values of cir-
culatory roadway widths than those of current methods. The difference ranges from 0.4 to
0.6 m for CITY-BUS, 0.7e1.0 m for WB-12, and 1.3e2.0 m for WB-20. The proposed guide-
lines would be useful in case of spatial restrictions.
© 2018 Periodical Offices of Chang'an University. Publishing services by Elsevier B.V. on
behalf of Owner. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

* Corresponding author. Tel.: þ1 416 979 5000.

E-mail addresses: [email protected] (M. Diachuk), [email protected] (S.M. Easa).
Peer review under responsibility of Periodical Offices of Chang'an University.
https://doi.org/10.1016/j.jtte.2018.01.006
2095-7564/© 2018 Periodical Offices of Chang'an University. Publishing services by Elsevier B.V. on behalf of Owner. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
362 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371

trailer combinations, while achieving adequate deflection for

1. Introduction small vehicles (FHWA, 2010). In such cases, a truck apron
may be used to provide additional traversable area around
Geometric parameters of roundabouts are governed by the the central island. However, truck aprons provide a lower
maneuvering requirements of the largest vehicles expected to level of operation than standard non-mountable islands and
travel through the roundabout. At roundabouts, the circula- are used only when there is no other means of providing
tory roadway width should accommodate the design vehicle adequate deflection for the design vehicle.
(Rodegerdts et al., 2010). Roundabout circulatory and entry Three design vehicles with long wheelbase are considered
widths are also important inputs for sight distance analysis in this study: (1) light two-unit truck (WB-12), (2) heavy tractor-
(Easa, 2017a, b). Appropriate vehicle-turning templates or semitrailer truck (WB-20), and (3) standard single-unit bus
computer-aided design (CAD)-based computer programs are (CITY-BUS). The basic proportions and overall dimensions of
normally used to determine the swept path of a design the vehicles are governed by AASHTO (2011) and are shown in
vehicle throughout the turning movements. Usually the left- Fig. 1. Obviously, WB-12 and CITY-BUS are most typical for
turn movement is the critical path for determining the urban roundabouts, while WB-20 is more critical for rural
circulatory roadway width. According to the AASHTO (2011), roundabouts. According to the North American standards
a minimum clearance of 0.6 m (2 ft) should be provided the vehicles' gross weights are taken as: 32.1 t for WB-12,
between the outside edge of the vehicle's tire track and the 40 t for WB-20, and 18 t for CITY-BUS (Infrastructure and
curb line. Gingrich and Waddell (2008) presented the Transportation Motor Carrier Division, 2016). Other
required circulatory width for various ICDs for different information needed for simulation includes gross weight
design vehicles. In some cases, particularly where the ICD is (Ontario Ministry of Transportation, 2013), gross weight
small or the design vehicle is large, the turning distribution among axles, curb weight (Wikipedia, 2016), and
requirements of a design vehicle may result in unsafely axle distribution (American Public Transportation
constricted circulatory roadway for passenger vehicles that Association, 2014). The weight is regulated by the vehicle
move behind on minimum radius. In such cases, the weights and dimensions guide (Saskatchewan Regulations,
circulatory roadway width can be reduced and a truck apron 2010; Road Safety Authority, 2013), which establishes the
provided outside of the central island mountable curb can be requirements for vehicle axle-load limits (FHWA, 2000).
used to accommodate larger vehicles.
The best known approach for swept path determination is
that of the society of automotive engineers (SAE) (Kizas, 1997).
Using this approach, the parameters of circular movements 3. Vehicle dynamics-based model
for single and multiple unit vehicles have been established.
This approach is static in nature and based mainly on The circulatory roadway width is determined using vehicle
geometric relations for a vehicle traveling along the curve swept path. To understand the phenomenon of swept path, it
path without considering the effect of vehicle speed. Thus, is essential to appreciate the significance of various parame-
currently the circulatory width is determined using vehicle ters affecting the turning behavior of a vehicle. The amount of
geometric interpretation either by the trigonometric swept path varies directly with the wheelbase length of a unit
functions of SAE method or using CAD procedures. The and inversely with a radius of the turn through which the
effect of vehicle dynamics behavior on the swept path at vehicle travels. The magnitude of swept path is also affected
roundabouts has not been considered in previous research. by the number and location of the articulation points on a
The purpose of this paper is to develop guidelines for vehicle, the length of the arc negotiated during the turn, and
roundabout circulatory and entry widths based on vehicle the speed of the vehicle.
dynamics that consider vehicle speed at the steadyesteady The following assumptions are acceptable for vehicle dy-
motion. The next sections describe the design vehicles and namics models: (a) a vehicle is assumed as plane dynamics
the vehicle dynamics-based model. Model validation and model, (b) the redistribution of longitudinal and transversal
design guidelines for circulatory and entry widths are then vertical reactions on vehicle wheels is neglected, and (c) the
presented, followed by the conclusions. modeling of contact forces between the tire and the road is
based on the simplified Pasejka's magic formula for pure tire's
slip (Bakker et al., 1989; Pacejka, 2002).

2. Design vehicles
3.1. Model description
An important factor in roundabout design is the need to
accommodate the largest motorized vehicle likely to use the Consider briefly the integration of motion equations, where an
roundabout. The turning path requirements of this vehicle, articulated vehicle is represented by two-dimensional (2D)
termed design vehicle, dictate the circulatory and entry dynamics model (Fig. 2). This dynamics model is different
widths of roundabouts. The choice of design vehicle varies from the traditional system equations used in the state-
depending on the approach roadway types and the sur- space of the bicycle model for multiple unit vehicles. The
rounding land use characteristics (Mills et al., 2011). In basic idea of this dynamics model is to account of the total
general, larger roundabouts are expected to accommodate number of tires' contact points with a road surface and the
large vehicles. However, in some cases, land constraints use of the matrix technique to represent model components.
may limit the ability to accommodate large trailer semi- The model uses a minimum amount of necessary
 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371 363

Fig. 1 e Design vehicles used in the study (AASHTO, 2011) (unit: m). (a) WB-20. (b) WB-12. (c) CITY-BUS.

information that is optimally suited for fast computations in semitrailer ST whose mass center is C'. The masses are
software packages such as MATLAB (Mathworks, 2017). located at distances eTR, eST relatively to the joint point P. In
According to the principle of theoretical mechanics, the process of angular motion the longitudinal axes of the
consider a two-unit vehicle moving in a global coordinate tractor and semitrailer form an articulation angle, j.
system XOY (Fig. 2) (Luijten, 2010; Rempel, 1999). The vehicle is The vehicle's position at any time is sufficiently charac-
represented by massless rods pivotally jointed at point P. The terized by the combination of radii vectors (RC and RC0 ) from
distributed masses mTR, mST and inertias ITR, IST are the origin of global coordinates to the unit mass centers.
concentrated at the physical centers of the units. The master However, it is more convenient to use the generalized co-
unit is a tractor TR whose mass center is C, connected by a ordinates by introducing the mobile sub-global coordinate

Fig. 2 e Schematic representation of articulated vehicle dynamics model.

364 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371

systems at the mass centers. The axes of the sub-global sys- consider the angular momentum change for each truck point
tems are parallel to the global system axes. The local coordi- Mk relative to the joint point P. Then,
nate systems xCy and x'C'y' (blue points in Fig. 2) are 8 X ! !  X !!ðeÞ !!
introduced within the sub-global coordinate systems and >
> PMk mTRk V TRk dt ¼ PMk F TRk þ PMk F P
>
>
can rotate around the centers C and C0 , respectively. The >
> k k |fflfflfflffl{zfflfflfflffl}
>
< !
equations of dynamics are composed relatively to sub-global 0 (4)
> X ! !  X ! ! ðeÞ ! !
coordinate systems (Gillespie, 1992; Popp and Schiehlen, >
> PM'k mSTk V STk dt ¼ PM'k F STk þ PM'k F'P
>
> k
>
> |fflfflfflfflffl{zfflfflfflfflffl}
1993). In addition, at any point M or M0 additional sub-local : k
!
coordinate systems could be input. For example, in the case 0
! !
of tires' points the systems are called wheels' coordinate where mTRk , mSTk , V TRk ; V STk are masses and velocities of point
systems (Pacejka, 2002), where the vectors of acting loads Мk of tractor and semitrailer, respectively.
! !
and forces are determined. Thus, the various degrees of The internal forces F P ; F'P at point P, from the tractor and
freedom of the dynamics model are combined to form the semitrailer, respectively, produce no moments, unlike the
vectors (column matrices), as follows external forces. Then, the generalized moments acting on the
0 1 0 1 0 1 0 1 tractor and semitrailer, respectively, are given by
XTR VTRX VTRx V_ TRx
B YTR C B VTRY C B VTRy C B V_ TR C 8 X! X !!ðeÞ
R¼B C B C B C B y C
@ fTR A; VG ¼ @ uTR A; V ¼ @ uTR A; J ¼ @ u_ TR A; >
< MTRP ¼
> MTRP ¼ PMk F TRk
fST uST uST u_ ST
k
X !
k
X !!ðeÞ (5)
>
>
: MSTP ¼ MSTP ¼ PMk F STk
0 1 0 ðeÞ 1 k k
uTR Fx
B uTR C B FðeÞ C Thus, the second couple of scalar equations of rotational
W¼B C B y C
@ uTR A; F ¼ @ A (1) dynamics has been formed. Through simplification and
MTRP
uST M transformation, the system of the second-order differential
STP
equations in vector-matrix format can be written as
where XTR is the movement of tractor's mass center along the
global axis OX, YTR is the movement of tractor's mass center J ¼ M1 ½F þ ðL þ TÞdiagðWÞV (6)
along the global axis OY, fTR is the tractor's rotational angle
Introducing auxiliary matrices for a more compact repre-
relatively to the vertical axis passing through the tractor's
sentation of Eq. (6), then
mass center, fST is the semitrailer's rotational angle relatively
       
to the vertical axis passing through its mass center, VG is the 0 0 1 0 0 0 0 1
Z2 ¼ ; E2U ¼ ; E2L ¼ ; D¼ (7)
velocity vector in the global coordinate system according to 0 0 0 0 0 1 0 0
the degrees of freedom, V is the vector of the main unknowns
   
in the tractor's local mobile coordinates, J is the vector that ðmTR þ mST Þ D  DT mST eTR E2U
L¼ (8)
characterizes the relative acceleration in the tractor's local mTR eTR E2U Z2
coordinate system, W is the vector of angular velocities uTR,
    
uST, F is the vector of the generalized forces acting in the di- Z2 D Z2 E2L
T ¼ mST eST cosðJÞ þ T sinðJÞ (9)
rections coinciding with the generalized coordinate system DT Z2 E2L eTR D
and reduced to the local coordinate system.
It is more convenient to describe the matrix of inertia co-
To obtain the motion equations, the fundamental theo-
efficients M as a sum of the constant part МI and the variable
rems of theoretical mechanics about the changes of mo-
trigonometric part МT
mentum and angular momentum are used. The impulse of the
!
two-mass model Q S is defined as M ¼ MI þ MT (10)
! ! ! ! !  
Q S ¼ Q TR þ Q ST ¼ mTR V TR þ mST V ST (2) ðmTR þ mST ÞðE2U þ E2L Þ mST eTR DT 
! ! MI ¼ (11)
where Q TR , Q ST are momentum vectors for the tractor and mTR eTR D ITR E2U þ E2L IST þ mST e2ST
! !
semitrailer, respectively, V TR , V ST are absolute velocity vec-     
tors for the tractor and semitrailer mass centers, respectively. Z2 D Z2 E2L
MT ¼ mST eST sinðJÞ þ T cosðJÞ (12)
The change of the mechanical system's momentum is
T
D Z2 E2L eTR D
caused by the action of external forces' resultant, as follows The transitions from the wheel, sub-local and local coor-
(Diachuk and Lykhodii, 2011) dinate systems to the global coordinate system (which have a
!  !  !  !  !  rotation angle x) are achieved through the following transition
dQ S dt ¼ dQ TR dt þ dQ ST dt ¼ mTR V TR dt þ mST V ST dt
matrix
X!ðeÞ X!ðeÞ
¼ F TRk þ F STk (3)
 
k k
cosðxÞ sinðxÞ
!ðeÞ !ðeÞ CðxÞ ¼ (13)
where F TRk , F STk are vectors of external forces acting at any k sinðxÞ cosðxÞ
points of tractor and semitrailer, respectively. To transfer the velocities of the tractor's local coordinate
The first couple of scalar equations are obtained as pro- system to the global coordinate system, the following matrix
jections of the vector equation (Eq. (3)) on the axes of tractor's is formed
local coordinate system. To account for the rotational motion,
 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371 365

different inscribed circle diameters (Gingrich and Waddell,

  2008). The values in this reference are presented in feet and
CðfTR Þ Z2
NðfTR Þ ¼ (14)
Z2 E2U þ E2L therefore they were converted to SI units for appropriate
diameter range by using linear interpolation. Thus, for WB-
Finally, using Eqs. (1)e(14), the displacements are obtained
20 the compatible diameters were: 40 m, 45 m, 50 m, 55 m,
using double integration, as follows
and 60 m. Given the calculated widths for the quasi-static
Z Z movement case, the errors in the simulated widths
V¼ Jdt; VG ¼ NðfTR ÞV; R ¼ VG dt (15)
relatively to the regulated widths were obtained,
respectively, as 0.6%, 0.5%, 1.5%, 0.7%, and 0.9%.
The integration of accelerations and velocities by Eqs. (14)
Similarly, for CITY-BUS, the respective relative errors are
and (15) is carried out numerically in MATLAB (Mathworks,
0.7%, 0.4%, 1.4%, 0.2%, 0.3%, and 0.5%. For WB-12 the
2017), using the solution method for stiff equation systems.
already calculated circulatory lane widths determined in the
Note that the power balance elements and simulation of
study of turbo-roundabout with spiral multi-lane design
tire-road contact (Uil, 2007) are not demonstrated here due
(Chan and Livingston, 2014) were used. In this research
to sufficient prevalence of information (Bakker et al., 1989;
report, the outer diameters were calculated as 33.08 m,
Pacejka, 2002; Schmid, 2011).
35.58 m, 40.82 m, and 49.96 m. This group corresponds to a
The conversion of the generalized vehicle models from
diameter range of 30e50 m with increments of 5 m used in
mathematics to animation is organized in several stages. First,
the present simulation. Subsequently, the simulated widths
according to vehicle geometric data an animation model is
were presented using interpolation on the grid from turbo-
created that visually reflects a final solution of dynamics
roundabout data. The errors of the simulated data were
equations and contributes to verifying the calculations. Sec-
determined as 2.5%, 1.1%, 0.8%, and 3.0%.
ond, the steering drive, a steering trapezoid in particular
Note that although for truck WB-20 the roundabout with
(Diachuk, 2007), is calculated and rationalized to provide
ICD diameter of 35 m usually goes with an apron to force the
accurate ratio of the steered wheels' turning angles
swept path reduction, Gingrich and Waddell (2008) in the
(Ackerman's angles). Third, the masses and axial loads are
table of Fig. 1 give the corresponding value of the
distributed based on the weighting parameters. Finally,
circulatory turning width. In order to compare the results
multiple calculations with any desired initial conditions are
during the model testing, the calculations for the condition
performed. Note that the dynamics model for a single-unit
of a truck circulating on a roundabout without an apron
vehicle is a special case of that presented in the dynamics
were made. Such conditions may correspond well to a
model of Eqs. (1)e(15), where one degree of freedom (slave
piecewise curved road section, to a 90 turn or to a U-turn
unit) is removed.
in a constricted space.

3.2. Model validation 3.2.3. Comparison with roadway width of geometric model
The differences between the values of the circulatory roadway
The presented vehicle dynamics model, which considers the width by SAE geometric model and by the dynamics model in
effect of circulatory speed, was validated before establishing quasi-static mode were calculated using the relative differ-
the design guidelines for the speed-based circulatory roadway ence 3m (%) which is given by
width. The validation was conducted by comparing the results
.
of the vehicle dynamics model for the case reflecting a quasi- 3m ¼ gl  gg gg  100 (16)
static mode and the results of existing static-based method.
where gg, g1 are widths based on the dynamics model of gross
Three types of comparison were made: comparison with
weight vehicle and the geometric model, respectively.
AASHTO articulation angle, comparison of roadway width
The direct comparison (Tables 1e3) shows almost identical
from other studies, and comparison with roadway width of
circulatory roadway widths obtained by means of proposed
geometric model.
vehicle dynamics method's special case (where the speed is
practically zero) and by the geometric model. This consistency
3.2.1. Comparison with AASHTO articulation angles
provides evidence of the validity of the proposed method.
According to AASHTO (2011), during the U-turn execution
In practice, the vehicles can move with different load levels
with maximum steering wheel rotation angle, the maximum
that lead to changes in the physical parameters and initial
articulation angle reaches a certain value. For WB-12 truck
conditions. However, the results of the simulation with
at the steering angle of 20.3 the articulation angle is equal
different truck weights show almost similar circulatory
to 46 , and for WB-20 truck at the maximum steering wheel
roadway widths regardless of vehicle load level. For this pur-
angle of 28.4 the articulation angle reaches 68.5 . The
pose the relative difference 3w is defined similar to Eq. (16)
simulations of this maneuver with specified conditions for
based on the simulated data of the dynamics model for the
truck models gave the articulation angles of 46.7 for WB-12
case of unloaded vehicle relatively to case of a fully-loaded
and 69.3 for WB-20, where the relative errors amounted
vehicle. In view of this, a comparison example is presented
1.5% and 1.2%, respectively.
in Fig. 3, showing the difference in the calculated width
when vehicles move on quasi-static mode with different
3.2.2. Comparison of roadway width from other studies
load levels. As noted, the small differences provide an
The circulatory roadway widths of the existing static-based
opportunity to constrict vehicle weight influence by means
method are presented for WB-20 truck and CITY-BUS for
366 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371

Table 1 e Results of estimated circulatory roadway width for WB-12a.

Parameter Inscribed circle diameter, ICD (m)
30 35 40 45 50 55 60 65 70 75 80
Calculated width from geometric model (m) 6.9 6.2 5.8 5.5 5.3 5.2 5.0 4.9 4.8 4.7 6.9
Simulated width value for gross 6.9 6.3 5.9 5.6 5.4 5.2 5.1 5.0 4.9 4.8 6.9
weight vehicle on quasi-static mode (m)
Maximum difference of simulated widths (m) 0.7 0.8 0.9 0.8 0.9 1.0 0.9 0.8 0.9 0.8 0.7
Speed range (km/h) 1e30 1e35 1e40 1e40 1e45 1e50 1e50 1e50 1e55 1e55 1e55
Coefficient a [m$(h/km)b, 104] 5.798 3.316 1.999 2.325 1.452 0.792 0.973 1.137 0.777 0.763 0.826
Coefficient b 2.090 2.185 2.275 2.189 2.282 2.410 2.325 2.258 2.335 2.315 2.275
Coefficient c (m) 6.895 6.284 5.882 5.602 5.381 5.205 5.072 4.961 4.863 4.782 4.714
R2 0.9999 0.9997 0.9995 0.9996 0.9994 0.9987 0.9993 0.9995 0.9993 0.9994 0.9995

Note: To convert from meter to feet multiply by 3.28.

a
Estimated model: g(v) ¼ avb þ c.

Table 2 e Results of estimated circulatory roadway width for WB-20a.

Parameter Inscribed circle diameter, ICD (m)b
35 40 45 50 55 60 65 70 75 80
Calculated width from geometric model (m) 14.1 11.7 10.2 9.3 8.6 8.1 7.7 7.4 7.1 6.8
Simulated width value for gross weight vehicle 13.9 11.6 10.2 9.3 8.6 8.1 7.7 7.4 7.1 6.8
on quasi-static mode (m)
Maximum difference of simulated widths (m) 1.3 1.4 1.7 1.4 1.6 1.9 1.7 1.9 1.7 2.0
Speed range (km/h) 1e35 1e40 1e45 1e45 1e50 1e55 1e55 1e60 1e60 1e65
Coefficient a [m$(h/km)b, 104] 20.38 5.58 2.67 2.67 1.625 0.99 1.19 0.72 0.86 0.45
Coefficient b 1.82 2.13 2.29 2.25 2.35 2.45 2.37 2.48 2.41 2.55
Coefficient c (m) 13.92 11.55 10.18 9.27 8.59 8.07 7.67 7.33 7.06 6.81
R2 0.999 1.000 0.999 0.999 0.999 0.998 0.999 0.998 0.999 0.998

Note: To convert from meter to feet multiply by 3.28.

a
Estimated model: g(v) ¼ avb þ c.
b
For ICD ¼ 30 m, the radius is too small and articulation angle is too large, causing the semitrailer to experience unstable jack-knife situation.

Table 3 e Results of estimated circulatory roadway width for CITY-BUSa.

Parameter Inscribed circle diameter, ICD (m)
30 35 40 45 50 55 60 65 70 75 80
Calculated width from geometric model (m) 6.0 5.7 5.4 5.2 5.0 4.9 4.8 4.7 4.7 4.6 4.6
Simulated width for gross weight vehicle on quasi-static mode (m) 6.1 5.8 5.5 5.3 5.1 5.0 4.9 4.8 4.7 4.7 4.6
Maximum difference of simulated widths (m) 0.4 0.5 0.6 0.7 0.8 0.7 0.8 1.0 0.8 0.7 0.6
Speed range (km/h) 1e30 1e35 1e40 1e45 1e50 1e50 1e55 1e60 1e60 1e60 1e60
Coefficient a [m$(h/km)b, 104] 2.34 1.21 0.81 0.42 0.20 0.33 0.17 0.04 0.11 0.24 0.36
Coefficient b 2.21 2.34 2.39 2.53 2.70 2.52 2.66 3.04 2.72 2.49 2.37
Coefficient c (m) 6.13 5.75 5.48 5.28 5.12 5.00 4.89 4.79 4.73 4.67 4.62
R2 1.000 0.999 0.998 0.998 0.997 0.997 0.997 0.992 0.995 0.998 0.999

Note: To convert from meter to feet multiply by 3.28.

a
Estimated model: g(v) ¼ avb þ c.

of model generalization, followed by propagation of the gross which consists in the offset between trajectories of two ve-
weight data for the particular case on the vehicle's entire load hicle's control points while turning. The low speed off-
range. tracking is mostly common when turning in a city on a small
radius. In low speed turns, the rear tires are pulled inward of
the steering path. The longer the vehicle's wheelbase or the
4. Guidelines for circulatory width tighter a turn, the more off-tracking occurs. High-speed off-
tracking is a speed-dependent phenomenon that results from
4.1. Difference between dynamic and static widths the tendency of the truck's rear to move outward due to the
lateral acceleration of the vehicle as it follows a curve at
In the literature, the phenomenon of the off-tracking is higher speeds. As the truck speed increases from very slow,
described (Bernard and Vanderploeg, 1980; Ervin et al., 1984), the off-tracking to the inside of the curve decreases until, at
 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371 367

(ICD), g is the width between curbs (face to face), g' is the

width required for bus, h is the low profile mountable curb
at apron, i is the inscribed circle, and j is the central island.

4.2. Regression models

To achieve both accuracy and simplicity, the relation between

the circulatory roadway width g and speed v (for each inscri-
Fig. 3 e Effect of truck weight: relative differences in bed circle diameter) is represented by a power function as
circulatory roadway widths between truck models with follows
maximum and minimum load degree for the quasi-static
mode. gðvÞ ¼ avb þ c (17)

where g is the circulatory roadway width (m), v is the circu-

some particular speed, the trailer rear axles follow exactly the latory speed (km/h), a, b, c are unknown parameters. The co-
tractor steering axle. efficients were estimated using MATLAB (Mathworks, 2017),
Consider the formation of circulatory roadway width on a and the results are shown in Tables 1e3. Fig. 5 shows the
roundabout for the three design vehicles (Fig. 4). The figure maximum difference in width depending on the ICD diameter.
shows, for each design vehicle, the simulated steady-state As an example, the best-fit model, based on Eq. (17), for the
movement at speed close to zero (quasi-static mode) and the case of WB-20 and ICD ¼ 60 m is shown in Fig. 6. The
movement on a subcritical speed that precedes vehicle circulatory roadway width is based on the fit-model
stability loss (Stevenson, 2004). For all vehicles on the quasi- parameters from Table 2. In general, the good convergence
static mode an important property is clear. That is, the of the power model can be seen, including the narrow
planes of steered wheels are simultaneously perpendicular confidence interval. The goodness-of-fit is very high as
to the rays emanating from a turning center (Ellis, 1969; indicated by the coefficient of determination (R2) which is
Pacejka, 2002). Moving with velocities appropriate to real very close to 1. Note that the computation of the function on
traffic, the effect from the inertial forces increases, leading
to diminishing the truck articulation angle either to the
truck units' offsets towards an external border (ICD curb).
The increasing of the lateral slip in tire-road contact spots
(Stoerkle, 2013) results offsets of perpendiculars dropped
from the instant turn center onto wheels' planes (Fig. 4,
position b). In case of trucks, the bias augmentation of the
tractor unit towards outside curb may be the consequence of
considerable resistance reactions from the semitrailer.
Other position designations in Fig. 4: a is the quasi-static
movement mode, b is the steady-state movement mode on
pre-critical speed, c is the remaining circulatory roadway Fig. 5 e Maximum differences between static and dynamic
width, d is the maximum width difference, e is the clearance circulatory roadway widths.
from wheels (0.61 m or 2 ft), f is the inscribed circle diameter

Fig. 4 e Formation of circulatory roadway width at a roundabout for different design vehicles at quasi-static and steady-
state dynamic modes. (a) WB-12. (b) WB-20. (c) CITY-BUS.
368 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371

explained by the linear zone of the tire's elastic deformation

which extends until the critical value (proportional range) of
the slip angle occurs. However, research has shown that the
range of the average passing speed could be 25e45 km/h
(Chan and Livingston, 2014).
Due to the larger potential sustainability of single-unit
vehicles compared with multiple-unit vehicles, the critical
speeds of CITY-BUS have larger values that lead to crossing
the curves for ICD equal to 65 m, 70 m, 75 m, and 80 m (Fig. 7).
Unlike the CITY-BUS curves, the curves for WB-12 and WB-20
Fig. 6 e Example of regression relation between circulatory are seeking to unite at a point due to the generalized critical
roadway width and speed for WB-20 truck at ICD ¼ 60 m. speed.
According to standard design a single-lane roundabout has
an apron with low profile curb. The critical case of roadway
zero speed is not feasible since each vehicle has a minimum use during circular motion enables truck's sweep on apron
sustained movement speed different from zero. provided a clearance e to the central island's curb is provided.
However, as seen on Fig. 4(a),(b), at high-speed circulating
4.3. Design charts mode in range of 25e35 km/h vehicle location is practically
provided in zone of remaining roadway c. Another situation
Using the regression parameters of Tables 1e3, design values is seen regarding SUV, where the longitudinal wheelbase is
for the circulatory roadway width were established as a determinative factor of circulatory width change.
function of circulatory speed for WB-12, WB-20, and CITY- As can be seen from Fig. 4(c), the apron is not used, because
BUS, as shown in Fig. 7. For each design vehicles, curves for the required width for single even large-dimension vehicle is
ICD ranging from 30 (in case of WB-20 from 35) up to 80 m beforehand less than the width required for any articulated
with an increment of 5 m are presented. The circulatory vehicle. To illustrate an example the roundabout diameters
speed ranges from near zero (1 km/h) that provides comparable to the minimum radii of vehicle turning were
simulation on quasi-static mode to the critical speed that is chosen, which are 35 and 30 m. Remark that it is possible to
limited by the conditions of vehicle motion stability. achieve a greater maximum difference between calculated
The curves of Fig. 7 allow making some conclusions. The width values if demand a bigger time value of transient
largest width differences occur for the case WB-20 and can process, but it would be contrary to the logic of real traffic
reach 1.5e2.0 m which is completely explained by the long- when maximum ¾ e1 of circle is used to change direction.
wheelbase of the semitrailer unit. The changes in the
circulatory width for WB-12 equal to 0.7e1.0 m, where the
tractor's short longitudinal wheelbase positively influences
the width decrease. However, the relatively short semitrailer 5. Guidelines for entry width
unit is not conducive to this. The changes in the circulatory
width for the case of CITY-BUS equal 0.45e0.95 m, which Entry width substantially affects the capacity of the round-
obviously is mainly provided by the large longitudinal about. The capacity of an approach does not depend merely
wheelbase. It is also noted that the curves have relatively on the number of entering lanes, but on the total width of the
low variability in the speed range of 1e15 km/h. This can be entry. For single-lane roundabouts, typical entry widths range

Fig. 7 e Design guidelines for roundabout circulatory roadway width at steady-state movement mode. (a) WB-12. (b) WB-20.
(c) CITY-BUS.
 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371 369

from 4.3 to 4.9 m. However, larger or smaller values than the Using precise maneuvering, the superposition of the inner
preceding range might be demanded based on speed re- truck trace and of the arc connecting roundabout legs is
quirements of the critical vehicle paths and site-specific modelled as shown by the red dashed line in Fig. 8. At this
design vehicle. moment, the trajectory of the inner track is equivalent to a
To model the roundabout entry width, the presented circular arc with entry radius. Thus, the minimum radius of
vehicle dynamics model combined with roundabout entry the inner point at entry zone within truck length was
elements allow the determination of the entry width of the defined. To find the external radius the farthest point of the
roundabout. Consider the geometry of roundabout entry for outer track from the same center is determined, knowing
WB-20 on a roundabout (Fig. 8). The following inputs and the coordinates of the conjugating curve center and the
conditions are assumed. The lane width in each straight positions of the external trace points. This is represented by
direction (approach width) is 4 m, the inscribed circle the lilac dashed-dot arc which limits the minimum adequate
diameter ICD is 60 m, the circulatory roadway width g is space for the truck's passage through the roundabout. The
8.1 m, which corresponds to the value determined earlier difference in the arcs' radii provides the required minimum
excluding the apron (Table 2), and the roundabout legs are width w.
connected by entry radius according to the standards The required entry roadway width was established for
(AASHTO, 2011). The vehicle motion mode is variable-speed: different design vehicles and various roundabout inscribed
decreasing of truck velocity in approach zone, deceleration circle diameters ICD as shown in Table 4. As noted, the
and yielding at the entry (if needed), and sequential required width ranges from 4.3 m to 5.4 m for WB-12,
transition from acceleration to the stationary mode at a 5.6 me7.6 m for WB-20, and 4.5 me5.1 m for CITY-BUS.
constant speed. Since an articulated vehicle occupies the Clearly, the requirements for WB-20 are greater than the
largest swept path at low speeds, the necessity of pre-stop at typical values suggested in geometric design guides. To
the entry zone was taken into account. The width path is investigate the effect of the approach width on the required
formed not just from the tracks of the tires' outer edges but entry width, two approach widths of 4 and 5 m were
increased by the value e of guaranteed clearance from the simulated. The required values of the entry roadway widths
curb on each side. were 5.95 and 5.90 m, respectively. These results show that
Obviously, the approach width of 4 m in the case of WB-20 the approach width has a negligible effect on the required
is critical; the swept path is formed within the rear inner (red entry roadway width. Thus, the values of Table 4 which are
dash line) and front outer (blue dash line) tracks (Fig. 8). At based on an approach width of 4 m can be used in design.
certain point the truck articulation angle changes the sign As can be seen, the more sloping arc occurs in the case of
and traces of the swept path exchange their locations larger approach width that requires less swept path's width
accordingly. Evidently, the smallest width at the entry area and less maneuver action.
is the distance between concentric arcs passing through the
track nearest inner point and farthest outer point within the
truck length.
6. Results and discussion

The results reveal that in some cases the required space for
accommodating a long-wheelbase vehicle can be reduced by
various methods and technical solutions. Here are some of
them:

1. The easiest way to reduce the swept path is to increase

smoothly the circulating speed, which will ensure the
centrifugal forces in the retraction of a semitrailer in the
radial direction and diminish the truck's folding angle.
However, this requires a certain professionalism of the
Fig. 8 e Illustration of maximum swept width definition at driver, which should be capable to keep the process of a
the entry to a roundabout. long-wheelbase vehicle's circulation in a stable state
without reaching the critical speed, when the tires' lateral
slip becomes uncontrollable.

Table 4 e Results of entry roadway width estimated using vehicle dynamics simulation model (approach width is 4 m).
Design vehicle Inscribed circle diameter, ICD (m)
30 35 40 45 50 55 60 65 70 75 80
WB-12 5.4 4.9 4.8 4.8 4.6 4.6 4.6 4.6 4.4 4.3 4.3
WB-20 e 7.6 6.9 6.6 6.2 6.1 6.0 6.0 5.7 5.7 5.6
CITY-BUS 5.1 5.0 4.9 4.8 4.7 4.6 4.6 4.6 4.6 4.6 4.5

Note: To convert from meter to feet multiply by 3.28.

370 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371

2. Attempts to reduce the swept path occupied by a truck and 1. The guidelines are based on a vehicle dynamics model that
to ensure the tracing accuracy on a roundabout have considers the effect of vehicle speed mode. The guidelines
already been realized by means of vehicle design solutions. were established by considering the largest value of the
An interesting technology to reduce circulatory width was tire-road adhesion coefficient. It should be noted, however,
offered by a German company (WABCO, 2017). The system that road surface and tire conditions are not always ideal.
intelligent OptiTurn function helps to increases the This can slightly change the width values. Obviously, tire
semitrailer turning radius up to 0.5 m, reducing the models reflect their typical specific parameters, such as
necessity of driver's care about steering accuracy on adhesion properties, pressure, and radial and lateral stiff-
small single-lane roundabouts, and making the ness. Therefore, the user should ensure that the parame-
semitrailer more maneuverable at short curves in city ters are consistent with local conditions.
conditions. This innovative function detects roundabouts 2. The results of this paper provide an important connection
and curves automatically and releases the pressure of the to the issues of real economy of road infrastructure design
pneumatic balloons in rear axle's suspension either lifts in the foreseeable future. There is a need to coordinate
the rear axle. This feature adds turning ability and makes some aspects of current/future transportation and auto-
a lesser swept path of the maneuver, increases tire life, motive technologies as elements of intelligent trans-
and makes the maneuver safer on smaller roundabouts. portation concept. The oncoming technologies of
Another possibility to reduce the circulatory width for autonomous trucks and online monitoring of current
multiple-unit trucks and to adjust it dynamically is the traffic flow will allow the dynamic redistribution of traffic
application of steer-by-wire technology (VSE, 2016), which in network and to reduce the workload on road in-
is not developed sufficiently in Canada and United States, tersections such as multi-lane roundabouts. Optimal
unlike in Europe where the use of euro-combi vehicles roundabout design providing both space and safety savings
becomes available even in cramped urban conditions on can be solved partly by means of a uniform redistribution
the overall road infrastructure. This technology provides of traffic flows in densely developed road networks.
automatic steering effect of semitrailer's wheels so that,
for example, wheels' tracks of steering axle and rear
semitrailer axle will be practically coincidental (Jujnovich
and Cebon, 2013). A variety of self-steering technology Conflicts of interest
can be used.
3. There are a number of emerging technologies that can help The authors do not have any conflict of interest with other
reduce the required circulatory and entry widths for trucks. entities or researchers.
For example, the programmable electronic brake system at
circular motion has the ability to adjust dynamically the
width by monitoring the articulation angle by activating
certain wheel brakes of the semitrailer. The data from
Acknowledgments
installed sensors in such system would be enough to create
The financial support of the Natural Sciences and Engineering
control algorithms. Moreover, such automatic control
Research Council of Canada (NSERC) is acknowledged. The
systems can prevent truck jack-knifing and semitrailer
authors are grateful to two anonymous reviewers for their
swinging, providing stability in the limited space.
thorough and most helpful comments.
4. It appears that the roundabout exit width will be
different from the entry width, considering the truck's
initial folded state. This would require a more compli- references
cated maneuver, larger swept path and exit width
correspondingly. Therefore, the roundabout entry and
exit widths should not be the same, indicating that the
AASHTO, 2011. A Policy on Geometric Design for Highways and
splitter island must be asymmetric. Future research Streets. American Association of State Highway and
should investigate the requirements for roundabout exit Transportation Officials (AASHTO), Washington DC.
width. In reality, the effect of width changes can be American Public Transportation Association, 2014. An Analysis of
slightly more significant than simulated values due to the Transit Bus Axle Weight Issues. MORR Transportation
redistribution of tire's vertical dynamic reactions, the Consulting Ltd., Winnipeg.
Bakker, E., Pacejka, H.B., Lidner, L., 1989. A new tire model with an
tire's lateral elasticity which increases sideslip, the mass
application in vehicle dynamics studies. SAE Technical Paper,
center's displacement, and the offset of transversal roll 890087.
center of vehicle units. Bernard, J.E., Vanderploeg, M., 1980. Static and dynamic
offtracking of articulated vehicles. SAE Technical Paper,
800157.
Chan, S., Livingston, R., 2014. Design Vehicle's Influence to the
Geometric Design of Turbo-roundabouts. Transoft Solutions
7. Conclusions
Inc., Richmond.
Diachuk, M., 2007. Optimization of design parameters of
This paper has presented revised guidelines for the round- commercial vehicles' steering trapezoid. Bulletin of the
about circulatory and entry widths of the roundabouts. Based Prydniprovs'k State Academy of Civil Engineering and
on this study, the following comments are offered. Architecture 8, 42e49.
 J. Traffic Transp. Eng. (Engl. Ed.) 2018; 5 (5): 361e371 371

Diachuk, M., Lykhodii, O., 2011. Evaluation of articulated vehicle Saskatchewan Regulations, 2010. Vehicle Weight and Dimension
maneuverability equipped by the semitrailer wheels' control Regulations. Regina, Saskatchewan.
system “steer-by-wire”. Journal of Sevastopol National Schmid, M., 2011. Tire Modeling for Multibody Dynamics
Technical University. Series: Machinery and Transport 143e145. Applications. University of Wisconsin-Madison, Madison.
Easa, S.M., 2017a. Design guidelines for symmetrical single-lane Stevenson, B., 2004. An Automated Vehicle Stability Control
roundabouts based on intersection sight distance. Journal of System for Heavy Earthmoving Articulated Vehicles
Transportation Engineering Part A: Systems 143 (10), 04017052. (Undergraduate thesis). Queensland University of
Easa, S.M., 2017b. Modeling of unsymmetrical single-lane Technology, Brisbane.
roundabouts based on stopping sight distance. KSCE Journal Stoerkle, J., 2013. Lateral Dynamics of Multiaxle Vehicles (Master
of Civil Engineering (in press). thesis). Swiss Federal Institute of Technology (ETH), Zurich.
Ellis, J.R., 1969. Vehicle Dynamics. Business Books, London. Uil, R.T., 2007. Tyre Models for Steady-state Vehicle Handling
Ervin, R.D., Fancher, P.S., Gillespie, T.D., 1984. An Overview of the Analysis (Master thesis). Eindhoven University of
Dynamic Performance Properties of Long Truck Combinations. Technology, Eindhoven.
UMTRI-84-26. University of Michigan Transportation Research VSE, 2016. Electronic Trailer Steering. Available at: www.v-s-e.
Institute, Ann Arbor. com/uploads/files/ets_trailer_eng_2009.pdf (Accessed 25
FHWA, 2000. Comprehensive Truck Size and Weight Study November 2016).
(TS&W). U.S. Department of Transportation. Federal WABCO, 2017. OptiTurn. Available at: www.wabco-auto.com/
Highway Administration, Washington DC. trailers/intelligent-trailer-program/trailer-safety/advanced-
FHWA, 2010. Roundabouts: An Informational Guide, second ed. safety/optiturn/ (Accessed 12 May 2017).
Transportation Research Board, Washington DC. Wikipedia, 2016. Curb Weight. Available at: www.en.wikipedia.
Gillespie, T.D., 1992. Fundamentals of Vehicle Dynamics. SAE org/wiki/Curb_weight (Accessed 15 December 2016).
International, Warrendale.
Gingrich, M., Waddell, E., 2008. Accommodating Trucks in Single
and Multilane Roundabouts. National Roundabout
Dr. Maksym Diachuk, post-doctoral fellow,
Conference, Kansas City.
is from Department of Civil Engineering,
Infrastructure and Transportation Motor Carrier Division, 2016.
Ryerson University, Toronto, Canada
Vehicle Weights and Dimensions Guide: Vehicle Weight and
(2016e2018). He was associate professor,
Dimension Limits in Manitoba. Available at: www.gov.mb.ca/
Department of Operation and Maintenance
mit/mcd/ (Accessed 13 December 2016).
of Machines, Mechanical Faculty, Prydni-
Jujnovich, B.A., Cebon, D., 2013. Path-following steering control
provs'ka State Academy of Civil Engineering
for articulated vehicles. Journal of Dynamic Systems,
and Architecture, Ukraine (1998e2016) and
Measurement and Control 135 (3), 031006.
associate professor, Department of Wheeled
Kizas, J., 1997. Design Vehicle Dimensions for Use in Geometric
and Tracked Transport Means, Faculty of
Design. Transportation Association of Canada, Ottawa.
Mechanical Engineering, National Metallur-
Luijten, M.F.J., 2010. Lateral Dynamic Behaviour of Articulated
gical Academy of Ukraine, Ukraine (2005e2014). In 2005, Dr.
Commercial Vehicles (Master thesis). Eindhoven University
Diachuk obtained his PhD in motor card and tractors, Kharkiv
of Technology, Eindhoven.
National Automobile and Highway University, Ukraine. His
Mathworks, 2017. MATLAB. Available at: www.mathworks.com/
research interests include automotive design and systems,
help/stateflow/ (Accessed 18 June 2017).
vehicle dynamics and control, solid mechanics, numerical
Mills, A., Duthie, J., Machemehl, R., et al., 2011. Texas Roundabout
methods in engineering, mathematical statistics and probability,
Guidelines. Center for Transportation Research, University of
simulation, and modeling.
Texas, Austin.
Ontario Ministry of Transportation, 2013. Highway Traffic Act.
Available at: http://www.e-laws.gov.on.ca/html/regs/english/
elaws_regs_050413_e.htm (Accessed 4 May 2016). Dr. Said M. Easa has been a professor of Civil
Pacejka, H.B., 2002. Tyre and Vehicle Dynamics. Butterworth- Engineering, Ryerson University, Canada
Heinemann, Oxford. since 2000. He received PhD degree from
Popp, K., Schiehlen, W., 1993. Ground Vehicle Dynamics. University of California at Berkeley and BSc
University of Stuttgart, Stuttgart. from Cairo University. He is fellow of Cana-
Rempel, M.R., 1999. Improving the Dynamic Performance of dian Academy of Engineering. He serves as
Multiply-articulated Vehicles (Master thesis). University of associated editor of several international
British Columbia, Vancouver. journals. Dr. Easa received numerous inter-
Road Safety Authority, 2013. Guidelines on Maximum Weights national, national, and local awards and
and Dimensions of Mechanically Propelled Vehicles and honors from Canadian and U.S. organiza-
Trailers, Including Manoeuvrability Criteria. Road Safety tions. His research interests include plan-
Authority, Primrose Hill. ning, design, operation, and management of transportation
Rodegerdts, L.A., Bansen, J., Tiesler, C., et al., 2010. Roundabouts: systems. He has introduced innovative approaches for safer and
An Informational Guide. National Cooperative Highway more efficient transportation infrastructure, and has identified
Research Program Report 672. Transportation Research ways for improving the design of highway alignments, railway
Board, Washington DC. crossings, interchanges, intersections, and roundabouts.